Band Pass Filter Design Part 3. Band Pass Filters From First Principles Richard Harris G3OTK

Download as pdf or txt
Download as pdf or txt
You are on page 1of 4
At a glance
Powered by AI
The document discusses using Norton's Transform to change impedances in filter circuits by replacing components with an equivalent circuit of capacitors. Norton's Transform can be used to increase impedances so that standard value inductors can be used or for impedance matching.

Norton's Transform is a technique for changing impedances within a filter. It can be used to equalize inductor values, change non-standard values to standard ones, or ensure impedance matching with a load. It demonstrates replacing a capacitor and transformer with an equivalent circuit of just three capacitors.

Norton's Transform shows that one of the replacement capacitors can have a negative value. While not physically realizable, the concept of a negative capacitor can still be used if combined with a larger positive capacitor. This allows obtaining a voltage step-up from just capacitors that replaces the transformer.

Band pass filter design

Part 3 - Using Nortons Transform


Richard Harris G3OTK
1. Introduction
Part 1 gave the procedure for design a third order filter from first principles. Sometimes it is
necessary to change the filter impedance part way through the filter, perhaps so that standard
values of inductors can be used or to give an insertion gain. One method is Nortons Transform
and Part 3 of this series demonstrates its use.
The filter developed in Part 1 Fig. 14 had the following specification:

Chebychev with 1dB pass band ripple (so that the ripple is visible on simulations)

Third order, capacitor input


Centre frequency (Fo) of 10MHz
Bandwidth (FBW) of 500kHz

The terminating resistance at the input and output was 2,540 Ohms if another terminating
resistance is required then a matching section must be used, and this will be discussed in Part 7
of this series.
The circuit diagram is shown in Fig 1.
2.54k

8.93p

8.93p

R1

C4

C5

Probe1-NODE

AC 1
V1

1u
L1244.4p
C1

2.54k
R2

1u
L3244.4p
C3

1u
L2235.4p
C2

Fig. .1. Filter circuit

dB(Probe1-NODE)-dB(Probe2-NODE) / dB

The insertion loss, as predicted by the SPICE simulator (Ref. 1), is shown in Fig. 2.
0
-5
-10
-15
-20
-25
-30
-35
-40

8.5

9.5

10

Frequency/MHertz

10.5

11

11.5

12

500kHertz/div

Fig. 2. Insertion loss

Version 1

Page 1

2. Nortons Transform
Nortons Transform (Refs. 2 & 3) is a useful technique for changing the impedance within a
filter. This might be necessary to equalise inductor values, or to change a non-standard
inductor value to a standard value or to ensure impedance matching with a load, such as the
input to an amplifier. The purpose of Part 3 of this series is to demonstrate this Transform. Part
4 will describe an alternative method.
On the left hand side of Fig. 9 is a capacitor connected in series with the primary winding of an
N-to-1 transformer. This combination can be replaced by three capacitors, as shown on the
right hand side. Of course the circuit on the right doesnt give the same isolation as that on the
left but that is not what we are trying to achieve.

(1 N) C

NC

N (N 1) C

N:1
Fig 3. Nortons Transform
So how is it possible to replace a capacitor and a transformer with just three capacitors, and,
perhaps more to the point, what if the transformer is a step-up transformer? How can you get a
higher voltage out of a combination of just three capacitors? The answer is quite simply that
one of the capacitors, either one on the left or the one on the right, will have a negative value.
We met negative capacitors in Part 1 and, whilst they are not physically realisable, the concept
can still be used, providing that there is a parallel capacitor that has a larger value so that the
combination is still positive. At a single frequency a negative capacitor can be replaced by an
inductor and this explains why it is possible to get a voltage step-up.
To demonstrate Nortons Transform, we will increase the impedance of the filter by a factor of
2.2 so that L3 is increased from 1uH to 2.2uH whilst keeping L1 and L2 at 1uH.
The
terminating resistance R2 must be increased and C3 must be reduced by the same factor.
Recall that transformers change impedance levels by the square of the number of turns, so the
turns ratio, N, of our notional transformer in Fig. 9 will be the square root of the ratio of L1 to L3,
which is 0.674.
Taking the right-hand side of the circuit diagram, as shown in Fig. 4.
8.93p
Probe1-NODE

C5
1u
L3

244.4p
C3

2.54k
R2

Fig 4. Right hand side of filter

Next add a transformer with a ratio of 0.674:1 between C5 and L3/C3/R2, whilst increasing the
value of L3 and R2 by a factor of 2.2 and reducing C3 by the same factor.

Version 1

Page 2

8.93p
Probe1-NODE
C5
TX1
P1

2.2u
L3

S1

111.1p
C3

5.588k
R2

Transformer
0.674 : 1
Fig. 5. Transformer added to increase impedance

Nortons Theorem enables us to replace C5 and TX1 by three capacitors.


6.02p
Probe1-NODE
C5

2.91p
C6

-1.96p
C7

2.2u
L3

111.1p
C3

5.588k
R2

Fig. 6. Replacing C5 and transformer with Nortons Transform capacitors

Now, although C7 is a negative value capacitor, it can be absorbed into C3, reducing it slightly.
C6 can be absorbed into C2 (see Fig. 1), increasing it slightly. So we have the final circuit
diagram for this filter
2.54k

8.93p

R1

C4

AC 1
V3

1u
L1

6.02p
Probe1-NODE

244.4p
C1

1u
L2

C5

238.3p
C2

2.2u
L3

109.1p
C3

5.588k
R2

Fig 7. Filter final circuit diagram

Version 1

Page 3

The insertion losses of the original filter in Fig. 1 and the modified filter in Fig. 7 can be
simulated and are shown in Fig. 8. The red curve is the response of the filter of Fig. 7. The
modified filter has an insertion gain because of the unequal terminating resistances. This gain
is the square root of the ratio of the output and input terminating resistances, 3.42 dB.
5
0
-5

dB

-10
-15
-20
-25
-30
-35
-40

8.5

9.5

10

10.5

Frequency/MHertz

11

11.5

12

500kHertz/div

Fig 8. Insertion losses of Fig. 1 and Fig. 7.

3. Conclusion

Nortons Transform can be used to increase the impedance within a filter so that standard
value, or specific value inductors can be used. The terminating resistances cannot be chosen
independently and the calculated values have to be used. However, matching sections can be
used to convert to some other terminating resistance as will be discussed in a later part to this
series.

4. References

1. www.simetrix.co.uk An excellent free fully functioning evaluation version of SPICE. The


limitation of this evaluation version is the number of components in the region of 100. This is
rarely a problem.
2. Anatol I. Zverev, Handbook of Filter Synthesis, Wiley Interscience. This book was first
published in 1967 and is regarded as a classic and is quoted as a reference in a great many
articles on filter design. It is now available new in paperback form.
3. Arthur B. Williams and Fred J, Taylor, Electronic Filter Design Handbook, McGraw-Hill.
This is a very good book, much more readable and usable than Zverev, at least for amateur
radio projects.

Richard Harris G3OTK


June 2010
[email protected]

R. J. Harris 2010

Version 1

Page 4

You might also like